Welcome to the Nexus of Ethics, Psychology, Morality, Philosophy and Health Care

Welcome to the nexus of ethics, psychology, morality, technology, health care, and philosophy

Monday, September 20, 2021

Intergroup preference, not dehumanization, explains social biases in emotion attribution

F. Enoch, S. P. Tipper, & H. Over
Volume 216, November 2021, 104865


Psychological models can only help improve intergroup relations if they accurately characterise the mechanisms underlying social biases. The claim that outgroups suffer dehumanization is near ubiquitous in the social sciences. We challenge the most prominent psychological model of dehumanization - infrahumanization theory - which holds outgroup members are subtly dehumanized by being denied human emotions. We examine the theory across seven intergroup contexts in thirteen pre-registered and highly powered experiments (N = 1690). We find outgroup members are not denied uniquely human emotions relative to ingroup members. Rather, they are ascribed prosocial emotions to a lesser extent but antisocial emotions to a greater extent. Apparent evidence for infrahumanization is better explained by ingroup preference, outgroup derogation and stereotyping. Infrahumanization theory may obscure more than it reveals about intergroup bias.


• Infrahumanization theory predicts outgroups are often denied uniquely human emotions.

• However, to date, antisocial uniquely human emotions have not been investigated.

• We test attributions of prosocial and antisocial emotions to social groups.

• Attributions of antisocial human emotions were stronger for outgroups than ingroups.

• We find no support for the predictions of infrahumanization theory.

From the General Discussion

Our results dovetail with recent empirical work that challenges the predictions made by Haslam's (2006) dual model of dehumanization (Enock et al., 2021). This research showed that when undesirable human-specific characteristics (such as ‘corrupt’ and ‘selfish’) are included in overall measures of humanness, there is no evidence for either animalistic or mechanistic dehumanization of outgroups as characterised by the dual model. Rather, desirable human qualities are more strongly attributed to ingroup members and undesirable human qualities to outgroup members. The present work extends these findings by further demonstrating the importance of considering sociality confounds when measuring psychological processes of ‘dehumanization’, this time through another highly prominent framework within the field.

During the review process, it was put to us that because dimensions of valence and sociality correlate highly in our pretest, the two constructs are “indistinguishable”, thus rendering our critique obsolete. We believe this represents a misunderstanding. Height and weight are strongly positively correlated, yet they are distinct constructs. Similarly, even though emotions that are generally perceived as prosocial may also perceived as positive to experience, and emotions that are generally perceived as antisocial may also be perceived as negative to experience, the two constructs are clearly conceptually distinct. While sadness is negative to experience, it is not inherently antisocial in character. Schadenfreude on the other hand is, by definition, positive to experience but antisocial in character.